The Robust Reliability of Neuropsychological Measures: Meta-Analyses of Test–Retest Correlations

Abstract

Test–retest reliability is an important psychometric property relevant to assessment instruments typically used in neuropsychological assessment. This review presents a quantitative summary of test–retest reliability coefficients for a variety of widely used neuropsychological measures. In general, the meta-analytic test–retest reliabilities of the test scores ranged from adequate to high (i.e., r=.7 and higher). Furthermore, the reliability values were largely robust across factors such as age, clinical diagnosis, and the use of alternate forms. The values for some of the memory and executive functioning scores were lower (i.e., less than r=.7). Some of the possible reasons for these lower values include ceiling effects, practice effects, and across time variability in cognitive abilities measured by those tests. In general, neuropsychologists who use these measures in their assessments can be encouraged by the magnitude of the majority of the meta-analytic test–retest correlations obtained.

Keywords:

• Reliability generalization
• Neuropsychological assessment
• Serial testing